Flame retardancy of natural fibre reinforced composites

Natural fibres offer various advantages over commonly applied man-made reinforcements, as lower density, renewability and biodegradability, however their low thermal stability and flammability represents a major drawback, especially in more demanding structural composite applications. The high amount of functionalizable groups on their surface offers various possibilities for rendering them FR (as summarized in 2.3.4). Nevertheless, it is still a challenging

0

2 mm DGEBA composite + 2x1 mm PER TEDAP coating 2 mm DGEBA composite + 2 mm PER TEDAP coating

127 task to be solved due the so called candlewick effect of natural fibres in polymer composites, as well as because of their low thermal stability and susceptibility to various chemicals.

In the followings results on FR modifications of twill woven hemp fabrics, as well as reference and flame retarded composites made thereof, will be presented.

4.5.2.1. Fire retardant modification of biofibres

The effects of thermotex procedure (i.e. removal of adsorbed water from the capillaries and then filling the micro/nano-voids with phosphoric acid [171]) and sol-gel treatment with amine-type silanes followed by thermotex treatment were compared on the flammability of hemp fibres used for the reinforcement of epoxy resins. The effect of the applied modifications (THF: thermotex-treated hemp fabric and SiTHF: silane and thermotex-thermotex-treated hemp fabric) on the thermal stability and flammability of the fibres, compared to the unmodified fabric (NHF: non-modified hemp fabric), was tested by TGA and mass loss calorimetry. The effect of surface modification on the tensile properties of the fabrics was evaluated by strip tensile tests [229].

Thermal stability, flammability and strip tensile strength of modified fabrics

The most important thermal, flammability and mechanical characteristics of the untreated and surface treated hemp fabrics are summarized in Table 4.5.18 (for detailed results see [229]).

Table 4.5.18 Thermal, flammability and mechanical characteristics of the modified hemp fabrics

fibre T_-5% temperature belonging to dTGmax; TTI: time to ignition; pHRR peak of heat release rate; THR: total heat release; Fmax: maximal force measured during the strip tensile test of the fabrics

Concerning the thermal stability, the thermotex procedure decreased the initial degradation temperature of the hemp fabric by more than 60 °C, as anticipated [229]. In the case of SiTHF fabric the sol-gel treatment partially protected the cellulose structure from the acidic hydrolysis [230,231], and thus increased the decomposition temperature by 30 °C compared to THF. The temperatures belonging to maximal mass loss rates showed similar tendency: the main degradation of the treated fabrics occurred at approx. 90° C and 70 °C lower temperatures in case of THF and SiTHF, respectively. On the other hand, the surface treatments considerably decreased the decomposition rates: in the case of the combined treatment the synergism between P and Si [163] resulted in a decrease from 1.71%/°C to 1.08%/°C. The surface treatment also increased the

128 amount of the residue significantly, but no remarkable difference (only 3.5%) was detected between the two treatments by TGA measurements.

In order to determine the macroscopic flammability of the fabrics, single fabric plies were subjected to mass loss calorimetric measurements. The application of the combined treatment was the most beneficial: it increased the time to ignition from 3 s to 15 s, decreased the heat release rate from 68 kW/m² to 9 kW/m² and led to the formation of consistent char instead of fluffy, light ash as in the case of untreated fabrics and thermotex treated fabrics.

According to the maximal forces measured during the strip tensile test of the fabrics, the strength of the treated fabrics decreased by about 35%. In this apect there was no significant difference between the two different surface treatments. These results also confirm that the acidic degradation of cellulose deteriorates the tensile properties of the natural fabrics.

4.5.2.2. Reactive flame retardancy of aliphatic epoxy resin based composites reinforced with flame retarded natural fibre

Reference and flame retarded composites were prepared from pentaerythritol-based epoxy resin (PER) and T58 curing agent with natural (NHF), thermotex treated (THF) and sol-gel and thermotex treated (SiTHF) twill woven hemp fabrics. In combination with the surface treated fabrics amine-type P-containing curing agent (TEDAP) was used as FR. The P-content of the flame retarded matrix was chosen for 2.5%, to reach V-0 rating in UL-94 tests, according to previous studies [173,220].

A possible synergistic effect was described in terms of composite mechanical properties: when both the matrix and the reinforcement contained P, despite the inferior mechanical performance of the FR matrix itself and the decreased strength of the surface treated fabrics, the mechanical properties for the FR samples reached the level of the reference composite almost in all cases.

Flame retardancy

The flame retardancy results of the hemp fibre reinforced composites are summarized in Table 4.5.19.

129 Table 4.5.19 LOI, UL-94 and cone calorimetry results of hemp fibre reinforced composites

sample P-content

LOI: limiting oxygen index, TTI: time to ignition, pHRR: peak of heat release rate, THR: total heat release

*in parenthesis the horizontal burning rate is showed in mm/min, where measurable

Due to the well-known candle-wick effect of the natural fibres [160] the application of untreated NHF reinforcement slightly reduced the LOI value of PER from 23 to 22 V/V%. This effect was compensated by flame retarded hemp fabrics: by applying THF the LOI increased by 4 V/V%

compared to the NHF reinforced composite, while when applying the combined treatment, an LOI value of 28 V/V% was reached. The application of NHF slowed the horizontal burning of the HB rated PER from 32 mm/min to 18.2 mm/min. When treated fabrics were used, no horizontal burning rates could be measured, however due to flaming up to the holding clamp, the samples did not reach the V-categories.

In TEDAP-containing matrices, the LOI value was increased by 3 V/V% by the use of reinforcement, independently from the surface treatment, which can be attributed to the intense charring of the FR. The effect of the P-Si synergism, observed in the case of the LOI values in PER-based non flame retarded composites, was overwhelmed by the effect of the FR curing agent. As a general rule, the use of at least 2% P is necessary to reach V-0 [103,104]. As the application of 30% untreated fabric as reinforcement decreased the overall P-content of the composite below 2%, so the UL-94 rating was only V-1 in PER TEDAP NHF composite. When treated fabrics were applied, the P-content of the composite increased and V-0 rating was reached again.

Considering the heat release rates, when flame retarded fabrics were applied in PER, pHRR was reduced by about 25%, compared to the NHF reinforced specimen, independently of the surface treatment.

The application of FR matrix decreased the pHRR values by about 32-42%, compared to the reference matrix composites. Comparing them to each other, the effect of the different surface treatments (THF and SiTHF) is more pronounced. In the case of THF, the pHRR appeared 15 s later than in the case of NHF, and its value decreased by 15%. Using SiTHF further 10% pHRR decrease was observed, while the maximum was reached 45 s later, compared to the NHF. The application of the treated fabrics decreased the THR of the reference matrix composites. In flame retarded

130 matrix composites no significant difference was found between the effects of the fabrics. The amount of the residues increased when treated fabrics were applied in the reference matrix. In the case of PER TEDAP composites, the highly charring character of the applied FR matrix overwhelmed the effect of the fabric treatment, therefore the quantity of the charred residue was almost the same for all samples, approx. 10 times higher than in case of PER NHF composite.

Mechanical characterization

The tensile, bending and interfacial shear strength test results of hemp fibre reinforced composites are presented in Table 4.5.20.

Table 4.5.20 Tensile, bending and interfacial shear strength test results of hemp fibre reinforced composites

PER NHF 66.50±3.21 3.26±0.36 92.79±10.11 4.91±0.64 13.21±2.20

PER THF 48.36±2.24 3.56±0.46 81.15±8.58 4.09±0.55 6.78±3.00

PER SiTHF 35.87±4.93 3.09±0.36 71.04±8.96 3.41±0.46 n. a.

PER TEDAP NHF 55.31±2.53 3.43±0.28 69.01±5.29 3.03±0.58 5.33±2.80

PER TEDAP THF 62.47±4.9 4.87±0.21 69.87±4.40 4.51±0.29 n. a.

PER TEDAP SiTHF 64.98±2.43 3.94±0.35 87.41±5.82 5.20±0.56 n. a.

As expected from the mechanical properties of the fabrics (see 4.5.2.1) the tensile strength of the composites decreased in PER reference composites, if the reinforcing fabric was surface treated.

Taking into account these results and the poorer mechanical properties of the FR matrix itself (see 4.4.5), it was unanticipated that the tensile strength of the FR composites with FR treated fabrics reached that of the PER NHF reference, not only in the case of the combined treatment, but also when thermotex treatment was applied alone. Similar trend was observed comparing the flexural strengths of the different composites, indicating increased fibre-matrix adhesion in composites, where both the matrix and the fibres were flame retarded.

Comparing the tensile modulus of the reference PER composites, no significant difference was observed between the different fabrics. When both the matrix and the reinforcing fabrics were flame retarded, slightly increased values were measured. Comparing the flexural modulus values, the PER reference composites showed decreasing values with the surface treatment, while in PER TEDAP composites the values increased when the fibres were flame retarded.

In order to explain the possible synergistic effect of flame retarding both the matrix and the natural fibres with P-containing FRs, the fibre matrix adhesion was determined by means of interfacial shear strength (IFSS) measurements applying microbond test (in the case of SiTHF fabric, elemental fibres could not be prepared as the fibres were attached to each other). Among

131 the measurable IFSS values, the highest one was measured in the case of untreated NHF fibres with PER reference matrix. In all other cases it decreased significantly. However, when both the matrix and the fibre were flame retarded, the method was not applicable because the droplets placed on the fibres were partially absorbed by the fibre, therefore the diameter of the spread droplets became too small to be caught by the blades of the microbond device. Consequently, no pull-outs could be detected at all. This phenomenon suggests considerably increased fibre-matrix adhesion, which explains the improved tensile and flexural properties of the composites consisting of matrix and natural fibres both flame retarded with P-containing species.